Below, we compute the first two terms of this sequence
explicitly. (When we compute
, we are doing the integration by
parts
,
. Since the integral runs from 0 to 1, the
boundary condition is 0 when evaluated at each of the endpoints. This
vanishing will be helpful when we do the integral in the general
case.)
![]() |
![]() |
|
![]() |
![]() |
|
![]() |
||
![]() |
||
![]() |
The reason that we defined this series now becomes apparent:
and that
. In general, it will be true
that
. We will now prove this fact.
It is clear that if the
were to satisfy the same recurrence that
the
and
do, in equation (5.3.1), then the above
statement holds by induction. (The initial conditions are correct, as
needed.) So we simplify
by integrating by parts twice in
succession:
![]() |
![]() |
|
![]() |
||
![]() |
||
![]() |
||
![]() |
||
![]() |
||
![]() |
||
![]() |
Therefore
. To conclude the proof, we consider the limit
as
approaches infinity:
by inspection, and therefore
Therefore, the ratio
William 2007-06-01